Generate a JSON representation of docs content

[cnunciato]

For easier consumption by internal services (e.g., Pulumi Copilot), we should generate a JSON document on every deployment containing all of our docs content, ideally in Markdown and indexed by URL path. For example:

[
    {
        "path": "/docs/get-started/index.html",
        "title": "Get Started with Pulumi | Pulumi Docs",
        "h1": "Get started with Pulumi",
        "main": {
            "innerHTML": "<p>Pulumi is an <a href=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/what-is/what-is-infrastructure-as-code></a>infrastructure as code platform [...]",
            "textContent": "Pulumi is an infrastructure as code platform [...]",
            "markdown": "Pulumi is an [](/what-is/what-is-infrastructure-as-code)infrastructure as code platform [...]"
        }
    },
    // ...
]

[cnunciato]

Fixed via #12004.